Technology ❯ Artificial Intelligence ❯ Cybersecurity

Threats

Vulnerabilities Automated Attacks Prompt Injection Exploitation Data Integrity Malicious Use Cyber Attacks Phishing Attacks Open AI Models Intellectual Property Global Impact AI in Special Ops 2 Financial Sector High-Performance AI Cyberterrorism AI-Driven Cybercrime Cybercriminals Prompt Injection Attacks Deepfakes Financial Institutions Model Extraction Autonomous Systems Banking Cybersecurity Defensive Strategies Hacking Mitigation Strategies Government Response Jewish Organizations Advanced AI Models Data Poisoning Distillation Attacks Safeguards Frontier AI Enterprise Risk Data Theft Agentic AI Critical Infrastructure Vishing AI Models Cyberattacks AI Agents Zero-Day Exploits Zero-Trust Architecture

OpenAI Ties Large-Scale Model Extraction Campaign to Moonshot AI Associates

OpenAI says the disclosure reveals a technical flaw that enabled scaled extraction of hidden model reasoning, prompting fresh industry and government action.

Linux Foundation Launches Akrites to Tackle Open-Source Security Flaws

New Tests Show Anthropic’s Mythos Can Automate Sophisticated Hacks as AI-Driven Intrusions Surge

Anthropic’s ‘Project Glass Wing’ Partners With CrowdStrike and Palo Alto, Lifting CRWD Shares

OpenAI Says Prompt Injection Is a Lasting Threat to AI Browsers, Rolls Out RL Red‑Teaming for Atlas

Mysten Labs Warns Bitcoin and Ethereum Need Quantum-Safe Upgrades

Anthropic-Led Study Finds About 250 Poisoned Documents Can Backdoor LLMs Regardless of Size